Normal spatial hearing is key to auditory scene analysis, which aids listeners in hearing out signals in the presence of competing sounds. Loss of this ability arguably is one of the greatest disabilities of people with mild-to-moderate hearin loss. We have characterized objective measures of scene analysis in human psychophysics and have examined some bottom-up mechanisms in auditory cortex in anesthetized cats. Now, we propose to examine the top-down cortical mechanisms of spatial aspects of auditory scene analysis, utilizing closely coordinated human and cat psychophysics with simultaneous cortical recordings in behaving cats. Specific Aim 1 addresses spatial stream segregation, by which listeners disentangle multiple interleaved sound sequences. In anesthetized conditions, cortical neurons exhibit the first steps of spatial stream segregation by synchronizing preferentially to one or the other of two competing sound sequences from differing source locations. Those anesthetized neurons, of course don't know which sequences correspond to target or to masker. Now, we will test the hypothesis that, during task performance with cortical recording, feline listeners select cortical modules synchronized to target sound sequences, either by facilitating modules synchronized to the target and/or by suppressing modules synchronized to the masker. Human and cat psychophysics and cat cortical physiology will evaluate spatial acuity in conditions in which listeners attend to spatial or to non-spatial segregation cues. Clinical reports and our recent results suggest that sound localization and spatial stream segregation are accomplished by different cortical areas. We will test this in humans by extending our observations that spatial segregation and localization differ markedly in their dependence on stimulus conditions. In cats, we will contrast spatial stream segregation among several candidate cortical areas. Aim 2 addresses spatial release from informational masking, which is the improvement in sound reception by spatial separation of a signal from a concurrent masker. Informational masking, in particular, is the masking that occurs in the absence of spectral overlap between signal and masker. In human and cat psychophysics, we will measure the trial-by-trial weights given to various masker components and will evaluate the degree to which various components and spatial cues contribute to spatial release. In cortical recordings from behaving cats, we will quantify masking of neural signal detection by out-of-band frequency components and the effects of spatial separation of signal and masker on rejection of such components. This work explores the monaural and binaural spatial cues for real-life auditory scene analysis. It will yield results that will inform design of sound processing strategies for hearing aids and cochlear implants. The new animal models that are introduced will yield new understanding of top-down task-dependent modulation of cortical spatial representation, enhancing diagnosis and treatment of spatial hearing deficits.